Purpose: Our objective was to evaluate the effect of viscosity in determining the cephalad spread of tetracaine administered by the subarachnoid route, Methods: We studied 42 patients that were randomly assigned to receive subarachnoid anaesthesia for arthroscope of the knee, They received 8 mg tetracaine dissolved in either glucose 10%, sodium chloride (NaCl) 5%, glucose 5%, or NaCl 2.5%, The specific gravity and viscosity of: each solvent was also measured. Cephalad spread was defined as the level of sensory block achieved, Results: The specific gravity of the glucose 5% and the NaCl 2.5% solutions were similar (1.0197 +/- 0.0001 vs 1.0177 +/- 0.0001 (mean +/- SD)). That was also true for the more concentrated pair of solutions (1.0385 +/- 0.0001 for glucose 10% vs 1.0353 +/- 0.0003 in NaCl 5%). However, the viscosities of the 10% and 5% solutions of glucose (0.01178 +/- 0.0002 and 0.01020 +/- 0.0002 (g . cm(-1)sec(-1))) were. higher(P < 0.05) than that of the other two solutions, being 0.00955 +/- 0.0001 for NaCl 5% and 0.00934 +/- 0.0002 for NaCl 2.5%. The median maximal extent of sensory block achieved was significantly higher(P < 0.05) with a solution of tetracaine in glucose 10% injected into the subarachnoid space ar the L3-4 interspace than that observed In the two groups administered tetracaine in NaCl. The level of sensory block 30 min after the injection of anaesthetic was; T-3 [T1-6] (median [range]) with glucose 10%, T-6 [T4-10] with NaCl 5%, T-5 [T2-11] with glucose 5%, and T-5.5[T2-11] with NaCl 2.5%. Conclusion: Viscosity was was shown to be one of determinants of the clinical extent of spread of a subarachnoid anaesthetic such as tetracaine.